7th International Conference on Spoken Language Processing

September 16-20, 2002
Denver, Colorado, USA

The 2ch Hybrid Subtractive Beamformer Applied to Line Sound Sources

Mitsunori Mizumachi, Satoshi Nakamura

ATR Spoken Language Translation Research Laboratories, Japan

A small-scale microphone array generally adopts the subtractive beamforming technique, which constructs sharp notches in a spatial beampattern, so that target sound sources are assumed as point sound sources. However, the shapes of actual sound sources are various and not always points. Our mouths are line sound sources or plane sound sources. It is important to deal with line sound sources when we apply the subtractive beamformer to speech signal processing. This paper proposes a hybrid subtractive beamformer that combines several 2ch subtractive beamformers properly in each frequency to expand the width of a sharp notch. The feasibility of the proposed hybrid subtractive beamformer is confirmed by performance tests in suppressing the signals from line sound sources.


Full Paper

Bibliographic reference.  Mizumachi, Mitsunori / Nakamura, Satoshi (2002): "The 2ch hybrid subtractive beamformer applied to line sound sources", In ICSLP-2002, 1833-1836.